Figures
Abstract
Background
VISION Max (Ortho-Clinical Diagnostics, Raritan, NJ, USA) is a newly introduced automated blood bank system. Cross-matching (XM) is an important test confirming safety by simulating reaction between packed Red Blood Cells (RBCs) and patient blood in vitro before transfusion. We assessed the benefits of VISION Max automated XM (A-XM) in comparison with those of manual XM (M-XM) by using multidimensional analysis (cost-effectiveness and quality improvement).
Materials and methods
In a total of 327 tests (130 patients), results from A-XM and M-XM were compared. We assessed the concordance rate, risk priority number (RPN), turnaround time, hands-on time, and the costs of both methods. We further simulated their annual effects based on 37,937 XM tests in 2018.
Results
The concordance rate between A-XM and M-XM was 97.9% (320/327, kappa = 0.83), and the seven discordant results were incompatible for transfusion in A-XM, while compatible for transfusion in M-XM. None of the results was incompatible for transfusion in A-XM, while compatible for transfusion in M-XM, meaning A-XM detect agglutination more sensitively and consequently provides a more safe result than M-XM. A-XM was estimated to have a 6.3-fold lower risk (229 vs. 1,435 RPN), shorter turnaround time (19.1 vs. 23.3 min, P < 0.0001), shorter hands-on time (1.1 vs. 5.3 min, P < 0.0001), and lower costs per single test than M-XM (1.44 vs. 2.70 USD). A-XM permitted annual savings of 46 million RPN, 15.1 months of daytime workers’ labor, and 47,042 USD compared with M-XM.
Citation: Chung H-J, Hur M, Choi SG, Lee H-K, Lee S, Kim H, et al. (2019) Benefits of VISION Max automated cross-matching in comparison with manual cross-matching: A multidimensional analysis. PLoS ONE 14(12): e0226477. https://doi.org/10.1371/journal.pone.0226477
Editor: Michael Nagler, Inselspital Universitatsspital Bern, SWITZERLAND
Received: May 31, 2019; Accepted: November 26, 2019; Published: December 23, 2019
Copyright: © 2019 Chung et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The importance of pre-transfusion tests, including cross-matching (XM), is the same as that of pre-transplantation laboratory tests; the importance of XM test, however, is easily underestimated because blood transfusions are routinely performed daily at the blood bank [1]. XM is an important pre-transfusion test confirming the compatibility of blood component for transfusion by observing the antigen-antibody reaction between blood component and patient blood in vitro [2,3]. If the patient is positive in unexpected antibody screening (ABS), the laboratory should identify the unexpected antibody so that they can issue compatible blood component for the patient when there is a transfusion order [3]. Electronic XM (also called computer XM) and automated XM are applied in some countries [4,5]. However, the policy of blood transfusion and blood supply varies greatly from country to country [4,5].
The result of the error in pre-transfusion tests can be critical or fatal [6,7]. Spillage or a small amount of the sample during XM tests may result in a re-examination and a delayed examination, and an error such as mislabeling of the patient sample may lead to an inadequate blood transfusion, even leading to patient death [7–9].
Recently, "patient safety" has been increasingly emphasized in healthcare, and efforts to prevent adverse events by reducing risk have been actively pursued [10–12]. From 2002, the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) began to mandate proactive risk assessment using failure mode and effects analysis (FMEA) to reduce the risk before an adverse event [13]. FMEA has been used in high-risk industries, such as in the astrospace sector and has been proved to be promising in reducing the risk of errors in the medical field [14,15]. In laboratory medicine, including transfusion medicine, the FMEA model is a useful tool in proactively analyzing and reducing the risks [16–21]. While reducing risk and reporting accurate results, it is also necessary to report the test results promptly to maintain the high quality of laboratory tests [22]. The FMEA model is already adopted in blood transfusion area to reduce risk and increase patient safety [16–18], and using the FMEA we previously reported benefits of automation in blood bank [17].
VISION Max (Ortho-Clinical Diagnostics, Raritan, NJ, USA) is a newly introduced automated blood bank system that is based on an agglutination method using a column containing a glass microbead matrix [23]. VISION Max automates the full range of immunohematologic testings, including ABO/Rh typing, XM, direct antiglobulin testing (DAT), ABS/antibody identification, and antigen testing [23,24]. Its middleware system is highly flexible and can be customized to each hospital's laboratory information system (LIS) [23,24].
Laboratory automation is an irreversible big trend [25–27]. Although automation of pre-transfusion testing processes can dramatically reduce error potentials and improve the safety of blood transfusion [28], clinical studies on automated XM (A-XM) is very limited [29,30]. In this study, we adopted VISION Max A-XM with customized middleware system and explored the benefits of using VISION Max A-XM in comparison with manual XM (M-XM) by multidimensional analysis. To assess performance, we observed the concordance rate; to assess quality, we observed the risk, turnaround time (TAT), and hands-on time; and to assess the effectiveness, we estimated costs. To the best of our knowledge, this is the first study on A-XM evaluation using VISION Max.
Materials and methods
Study design
This study was conducted in Konkuk University Medical Center (KUMC), a 700-bed tertiary-care teaching hospital, according to the Declaration of Helsinki. The protocol was approved by the Institutional Review Board of KUMC (KUH12000111), with an exemption of obtaining informed consent from the participants [31]. This reporting followed the STARD guidelines [32]. A-XM and M-XM were performed simultaneously with split-sample, and both results cannot be referred to the medical technician. After user education, training, and 10 weeks of familiarization period to provide sufficient experience to the medical technician, we enrolled all the consecutive clinical samples (regardless of age, gender, department, and patient setting) that were sent to the blood bank for routine XM testing between 2 pm and 4 pm from March 2 to March 31, 2018. Enrolled sample group in this experiment was exactly the same setting with the whole sample group undergoing XM test in the clinical setting. Because the subject of this paper is a sample, not a patient, sample eligibility followed the specimen rejection criteria of laboratory standard operation manual for XM testing of KUMC (S1 File). Laboratory step by step protocols are shared in protocol.io (dx.doi.org/10.17504/protocols.io.652hg8e) according to the instruction for the author. Specimen rejection criteria were: patient sample older than 24 hours from specimen collection, patient sample without appropriate labeling (patient’s identification number in hospital/name/age/gender and name of phlebotomist), hemolyzed sample by visual inspection, and not enough specimen less than 1.5 mL as serum/plasma (S1 File). Commercially available BioVue Screen Panels (Ortho-clinical Diagnostics, Raritan, New Jersey, USA) was used for ABS. Antibody identification was done in ABS positive samples, using 0.8% ORTHO RESOLVE Panel (Ortho-clinical Diagnostics) which is composed of 11 cells.
VISION Max provides flexible middleware, which bi-directionally communicates with the LIS and the analyzer [23,24]. The middleware is customizable to the situation of each hospital and blood bank, using tailored reflex test, automated verification, and responsive automation. VISION Max automatically accumulates all the compatible/incompatible XM results and provides statistics. For discordant results, results of ABO/Rh typing, ABS/antibody identification, DAT, and history of blood transfusions within three months were reviewed. We further simulated their annual effects based on the test numbers in 2018.
Cross-matching and concordance rate
A total of 327 samples (from 130 patients) were used to compare the results between A-XM and M-XM. The necessary sample size was 73 which was calculated based upon the accumulated incompatible rate 5.0% of our laboratory (95% confidence level, 0.05% of margin of error) by Cochran’s sample size recommendation [33]. Venous whole blood (6 mL) was drawn into BD Vacutainer serum tube (Becton Dickinson and Company, Franklin Lakes, NJ, USA), and serum was separated by centrifuging at 1000 g for 10 mins in room temperature.
A-XM was performed in VISION Max, using a Poly Cassette (Ortho-Clinical Diagnostics) including anti-IgG and anti-C3d. Briefly, 500 uL of donor red blood cells (RBCs) is added to the test tube, and the test tube and the plasma of the recipient are placed into the rack and then placed in the VISION Max. The barcode of the blood component is scanned. Thereafter, the robotic pipetting arm inside VISION Max automatically transfers 50 uL of BLISS, 40 uL of plasma, and 10 uL of 3% RBCs to Poly Cassette. The Poly Cassette is incubated at 37°C for 10 minutes and centrifuged for 5 minutes. Results of agglutination are determined as 0, ±, +1, +2, +3, and +4. The camera inside VISION Max reads the column and records agglutination grade with a digital image. Negative results of A-XM are assigned as compatible, and all the other results are assigned as incompatible. In a batch, seven XM tests can be performed simultaneously.
M-XM was sequentially tested through the 1st saline phase, 2nd albumin phase, and 3rd anti-human globulin phase. The M-XM process was performed according to the laboratory standard operation manual of KUMC (S1 File) based on Technical Manual of American Association of Blood Banks [2]. All XM tests were performed by the same experienced medical technician and were determined as compatible or incompatible. The concordance rate was calculated by the proportion of concordant pairs over the total number of paired tests. [34]. The concordance rate of A-XM and M-XM results was assessed overall and according to the presence of unexpected Ab, that is ABS positivity. Because unexpected Ab in patient blood more often produces incompatible results in XM with donor blood component. Therefore the authors expected that the concordance rate of A-XM and M-XM results will be higher [3].
Risk assessment by FMEA
According to the international standard for FMEA, the risk priority number (RPN) was calculated to quantify risks as following formula: RPN = severity (S) × occurrence (O) × detection (D) [35]. All three medical technicians working at the blood bank, one attending medical doctor, and one quality manager participated in the FMEA and RPN scoring [35]. During A-XM and M-XM processes, each step of workflow was separated, described, and reviewed. In each step, possible failure modes and subsequent potential effects were described based on the laboratory logbook, troubleshooting records, and interviews based on our previous FMEA assessment study in transfusion area [17] and International Standard for FMEA [35]. Based on this, rating score and operational definitions of severity of failure, frequency of occurrence, and detection of failure were determined (Table 1) [17]. As an example of RPN scoring, mis-recording of M-XM result happens with a frequency of less than once a year in our laboratory. In this event, severity score is 10 and the occurrence score is 1 and because all the M-XM results are cross-checked by a different medical technician, the detection score is 7, according to Table 1. So, the calculated RPN score of this event is severity (10) × occurrence (1) × detection (7) = 70.
TAT and hands-on Time
The TAT was defined as the time from the point at which XM testing was started to the time point at which the verified result was reported to the hospital LIS. To measure TAT, XM testing process was performed seven times and was recorded as a video; the video records were analyzed in seconds for TAT in each process step. Hands-on time was defined as the time taken by medical technicians for XM testing; it was calculated by subtracting the sample processing time in the Vision Max from the TAT.
Cost-effectiveness
The total costs included direct and indirect costs. Direct costs included costs for reagents and consumables per single test in A-XM and M-XM. Instrument cost was included in the direct cost as lease contract. Indirect costs included depreciation costs and labor costs. Depreciation costs were taken into account only when the instruments were directly related to the tests and purchased within the last five years [36]; as any of the instruments in the present study were not purchased within the last five years, depreciation costs were not included. Labor costs were calculated on the basis of the average operator salaries and hands-on time. Labor costs may vary across institutions; therefore, it was calculated by referring to the published multi-center research paper in Korea, considering 3.0% of the annual domestic inflation rate [36].
Statistical analyses
The Cohen's kappa was used to assess agreement between the two categories [37]. Homogeneity of variance was tested by performing F-test and Wilcoxon rank-sum test was used to evaluate the differences of TAT and hands-on time between A-XM and M-XM processes [34]. The data with non-parametric distribution were presented as median and interquartile range [38]. RPNs between A-XM and M-XM processes were relatively compared because RPN has no standard. The level of significance for all statistical analyses was set to P < 0.05. Statistical analyses were performed using Microsoft Excel 2016 (Microsoft Corp., Redmond, WA, USA) and R-3.3.2 under CentOS Linux 7.
Results
Concordance rate between A-XM and M-XM
The concordance rate between A-XM and M-XM was 97.9% (320/327, kappa = 0.85), showing near perfect agreement (Table 2). The seven discordant results were all from different patients. All the seven cases showed incompatible for transfusion in A-XM testing, while showed compatible for transfusion in M-XM testing. None of the results was incompatible for transfusion in A-XM while compatible for transfusion in M-XM, meaning A-XM detect agglutination more sensitively and consequently provides at least equally safe results with M-XM. Grades of incompatible results with A-XM are shown in S1 Table. Among the seven discordant results, three cases (43%) had a transfusion history within 90 days. All the seven discordant cases were negative for auto-antibody. DAT was requested only in two cases, and both of them showed positive results. The ABS positivity was 29% (2/7), having anti-Lea antibody and anti-P1 antibody, respectively. Anti-Lea antibody and anti-P1 antibody are IgM antibodies rarely the destruction of red blood cells even though presence of those specific antigens. Consequently rarely causing hemolytic transfusion reaction or hemolytic disease of the newborn. We did not evaluate the presence of those specific antigens, respectively. The discordant rate in the ABS-positive group was three times higher than that in the ABS-negative group (4.5% [2/44] vs. 1.8% [5/283]), as expected. In ABS positive group, 14 cases showed incompatible results in both A-XM and M-XM. Their antibody specifications were as follows: anti-E (4 cases), anti-M (3 cases), anti-Lea (2 cases), anti-Leb (2 cases), anti-P1 (1 case), and unidentifiable antibody (2 cases).
Discordant results have shown in bold.
Comparison of quality and cost between A-XM and M-XM
The A-XM and M-XM processes were determined to have 8 and 18 steps, respectively (Table 3). A-XM was estimated to have a 6.3-fold lower risk (229 vs. 1,435 RPN score) per single test than M-XM. A-XM permitted an annual reduction of 21 million RPN compared with M-XM (Table 4). A-XM was estimated to have a shorter TAT (19.1 vs. 23.3 min, P < 0.0001) and shorter hands-on time (1.1 vs. 5.3 min, P < 0.0001) per single test than M-XM. A-XM permitted an annual saving of 15.1 months of daytime workers’ labor (2,656 hrs of hands-on time) compared with M-XM (Table 4).
Regarding cost, A-XM was estimated to have lower costs per single test than M-XM (1.44 vs. 2.68 USD) (Tables 4 and 5). A-XM had higher direct costs (1.02 vs. 0.67 USD) but lower indirect costs (0.42 vs. 2.03 USD) than M-XM. It was estimated that A-XM could save 47,042 USD annually compared with M-XM, with saved indirect cost (60,547 USD) and increased direct cost (13,505 USD). The calculated average operator salary per minute in the present study was 0.38 USD (429 Korean won) [36].
Discussion
To the best of our knowledge, this study is the first to implement A-XM using VISION Max in a clinical transfusion laboratory. Laboratory automation is now expanding its reach into hematology, urinalysis [25–27], microbiology [39,40], special immunology [41], and even into blood banks [29,30]. While there is rapid implementation of automated blood grouping in clinical laboratories, automation of XM in clinical laboratories is in a very early stage [29,30]. The benefits of the A-XM using VISION Max, as seen in our results, were reduced risk, reduced TAT and hands-on time, and reduced cost through test automation.
In addition to the direct benefit related to automation, the other biggest advantage of the VISION Max system is that its middleware system connects LIS and VISION Max bi-directionally; the middleware can review the previous results and other related orders of a patient and suggest a reflex test when necessary. In our study, all seven discordant cases were A-XM incompatible and M-XM compatible, suggesting that the A-XM can determine agglutination as positive more sensitively than visual inspection. The clinical significance of the additional positive A-XM results remains open. We expect the increase in safety will be achieved by adopting A-XM as shown in the risk assessment in this study. Based on the discordant results, the pre-transfusion workflow using VISION Max and middleware system was proposed (Fig 1). Since we start to adopt A-XM as first-line routine testing in clinical laboratories, we considered the introduction of A-XM conservatively and sequentially for patient safety. It was based on the comparative analyses and customized rules in the blood bank of KUMC. And in the presence of any factors that may affect RBC antigen-antibody immunologic response, we made a record of additional double-check by M-XM (as repeat test). Using the reflex test function of middleware, the A-XM is actively used as a routine laboratory practice. In the review of the data for 8 months after the introduction of A-XM workflow (Fig 1), there has been no discrepant result between A-XM and M-XM. We reviewed the additional specification of the incompatible result of A-XM after introduction into the routine. For eight months, a total of 7,235 A-XM were tested and among them, incompatible results were 5.00% (362/7,235). Among incompatible 362 results, ABS positivity was 57.2% (207/362). After accumulating more experience, we can adopt more efficient rules and reflex tests. However, even after implementing A-XM, M-XM is still being performed in a small number of samples as a comparative method when necessary or as an alternative in an emergency setting. If multiple units of packed RBCs are requested, performing both A-XM and M-XM simultaneously is more efficient than waiting for sequential results of A-XM; in an emergency, abbreviated M-XM can be performed, and then packed RBCs can be issued [1,3].
It was based on the comparative analyses and customized rules for the blood bank of Konkuk University Medical Center. Abbreviations: ABS, antibody screening test; DAT, direct antiglobulin test; A-XM, automated cross-matching; M-XM, manual cross-matching.
In this study, cost reduction was a noticeable, unexpected advantage. Before the introduction of VISION Max, there were some concerns regarding the cost increase. Actually, the individual cost of consumables is relatively higher in A-XM than in M-XM. Increase of direct costs, including consumables and reagents, was 0.35 USD/test costs (from 0.67 USD/test to 1.02 USD/test). However, the reduction in indirect costs was 1.59 USD/test due to the reduction of the medical technician’s hands-on time, which was much higher than the increase in direct costs (from 2.03 USD/test to 0.42 USD/test). The procedure for the additional positive results in A-XM should also be included for the assessment of the costs and the time required (repetition with M-XM, other packed red blood cells tested). But this was an irregular event and a very small proportion, we did not concern in comparison of cost and time. Contrary to the initial concerns, the total costs as well as hands-on time were reduced. By reducing the hands-on time spent on the XM testing, medical technicians could have more time to spend in improving the test quality or in performing other tests.
Another advantage of VISION Max system is data accumulation and easy statistics. If one XM testing is ordered, a medical technician performs the XM testing until he/she finds a compatible/least incompatible blood component. In M-XM, medical technicians run the following XM continuously leaving only a transient manual record, if the results are incompatible; therefore, it was impossible to identify the actual number of XM testing performed and to accumulate data in M-XM. However, for A-XM using VISION Max, the compatibility of each XM Poly Cassette is stored as an image with agglutination grade on a separate laboratory server from the hospital LIS, functioning as a backup. After the introduction of VISION Max at the blood bank of KUMC, the incompatible result of A-XM was 6.7% (199/2,954 tests) for two months from 1 October to 31 December, 2018.
This study has some limitations. In A-XM, only the 3rd phase (anti-human globulin phase) was performed, whereas in M-XM three phases were performed sequentially. Since 1984, the American Association of Blood Banks recommended that the full XM, including anti-human globulin phase, could be replaced by an abbreviated XM in patients with negative ABS [1,3]. According to each laboratory’s policy, abbreviated XM is performed worldwide [4,42–44]. If the result in each phase is inconsistent in full XM, results of the 3rd phase determines compatibility in most cases [2]. We assumed that a different result in test phase may not have a big impact in practice. Additionally, in TAT analyses, time measurement was based on only seven XM tests, which is generally not a sufficient number for comparison; however, the TAT results were very similar with narrow ranges. Another limitation is that, considering the variation of laboratory and pre-transfusion testing across institutions, our estimation on quality and cost may not be extrapolated directly to other laboratories but can be used as one of the references. Further studies are needed for the blood bank automation including A-XM. In conclusion, this is the first attempt to implement A-XM using VISION Max in a clinical laboratory. Our multidimensional analysis showed that VISION Max A-XM is a safe, practical, and reliable alternative for pre-transfusion workflow with the potential to improve quality and cost-effectiveness in blood banks.
Supporting information
S1 Table. Description of the seven cases showing discordant results between M-XM and A-XM.
https://doi.org/10.1371/journal.pone.0226477.s001
(DOCX)
S1 File. Standard operating procedure for XM in blood bank of Konkuk University Medical Center.
https://doi.org/10.1371/journal.pone.0226477.s002
(DOCX)
Acknowledgments
The authors acknowledge the Ortho Clinical Diagnostics and medical technicians of department of laboratory medicine, KUMC, for the technical support.
References
- 1. British Committee for Standards in Hematology, Milkins C, Berryman J, Cantwell C, Elliott C, Haggas R, et al. Guidelines for pre-transfusion compatibility procedures in blood transfusion laboratories. British Committee for Standards in Haematology. Transfus Med. 2013;23: 3–35. pmid:23216974
- 2.
Harm SK, Dunbar NM. Transfusion-service-related activities: Pretransfusion testing and storage, monitoring, processing, distribution, and inventory management of blood components. In: Fung MK, Eder AF, Spitalnik SL, et al. eds. Technical manual. 19th ed. Bethesda, MD: American Association of Blood Banks; 2017:457–87.
- 3.
American Association of Blood Banks Standards Program Committee. Standards for blood banks and transfusion services. 31st ed. Bethesda, MD: American Association of Blood Banks; 2012.
- 4. Nasr IH, Papineni McIntosh A, Hussain K, Fardy MJ. Preoperative cross-matching in major head and neck surgery: A study of a department's current practice and eligibility for electronic cross-matching. Oral Surg Oral Med Oral Pathol Oral Radiol. 2013;116: 534–539. pmid:24021773
- 5. Kuriyan M, Fox E. Pretransfusion testing without serologic crossmatch: approaches to ensure patient safety. Vox Sang. 2000;78: 113–118. pmid:10765147
- 6. Sidhu M, Meenia R, Akhter N, Sawhney V, Irm Y. Report on errors in pretransfusion testing from a tertiary care center: A step toward transfusion safety. Asian J Transfus Sci. 2016;10: 48–52. pmid:27011670
- 7. Maskens C, Downie H, Wendt A, Lima A, Merkley L, Lin Y, et al. Hospital-based transfusion error tracking from 2005 to 2010: Identifying the key errors threatening patient transfusion safety. Transfusion. 2014;54: 66–73. pmid:23672511
- 8. Ansari S, Szallasi A. 'Wrong blood in tube': Solutions for a persistent problem. Vox Sang. 2011;100: 298–302. pmid:20738838
- 9. Dzik WH, Murphy MF, Andreu G, Heddle N, Hogman C, Kekomaki R, et al. An international study of the performance of sample collection from patients. Vox Sang. 2003;85: 40–47. pmid:12823729
- 10. Schwendimann R, Blatter C, Dhaini S, Simon M, Ausserhofer D. The occurrence, types, consequences and preventability of in-hospital adverse events-a scoping review. BMC Health Serv Res. 2018;18: 521. pmid:29973258
- 11. Makary MA, Daniel M. Medical error-the third leading cause of death in the US. BMJ. 2016;353: i2139. pmid:27143499
- 12. Miligy DA. Laboratory errors and patient safety. Int J Health Care Qual Assur. 2015;28: 2–10. pmid:26308398
- 13.
Joint Commission on Accreditation of Healthcare Organizations. Comprehensive accreditation manual for hospitals. Oakbrook Terrace, IL: Joint Commission on Accreditation of Healthcare Organizations; 2003.
- 14. Krouwer JS. An improved failure mode effects analysis for hospitals. Arch Pathol Lab Med. 2004;128: 663–667. pmid:15163233
- 15. Coles G, Fuller B, Nordquist K, Kongslie A. Using failure mode effects and criticality analysis for high-risk processes at three community hospitals. Jt Comm J Qual Patient Saf. 2005;31: 132–140. pmid:15828596
- 16. Chiozza ML, Ponzetti C. FMEA: A model for reducing medical errors. Clin Chim Acta. 2009;404: 75–78. pmid:19298799
- 17. Han TH, Kim MJ, Kim S, Kim HO, Lee MA, Choi JS, et al. The role of failure modes and effects analysis in showing the benefits of automation in the blood bank. Transfusion. 2013;53: 1077–1082. pmid:23002928
- 18. Lu Y, Teng F, Zhou J, Wen A, Bi Y. Failure mode and effect analysis in blood transfusion: a proactive tool to reduce risks. Transfusion. 2013;53: 3080–3087. pmid:23560475
- 19. Saxena S, Kempf R, Wilcox S, Shulman IA, Wong L, Cunningham G, et al. Critical laboratory value notification: a failure mode effects and criticality analysis. Jt Comm J Qual Patient Saf. 2005;31: 495–506. pmid:16255327
- 20. Jiang Y, Jiang H, Ding S, Liu Q. Application of failure mode and effects analysis in a clinical chemistry laboratory. Clin Chim Acta. 2015;448: 80–85. pmid:26116892
- 21. Mora A, Ayala L, Bielza R, Ataulfo Gonzalez F, Villegas A. Improving safety in blood transfusion using failure mode and effect analysis. Transfusion. 2019;59: 516–523. pmid:30609064
- 22. Shahangian S, Snyder SR. Laboratory medicine quality indicators: A review of the literature. Am J Clin Pathol. 2009;131: 418–431. pmid:19228647
- 23. Park Y, Kim SY, Koo SH, Lim J, Kim JM, Lim YA, et al. Evaluation of the automated blood bank systems IH-500 and VISION Max for ABO-RhD blood typing and unexpected antibody screening. Lab Med Online. 2017;7: 170–175.
- 24.
Product manual of Vision Max automated system. [cited 2019 May 20]. Available from:https://www.orthoclinicaldiagnostics.com/en-us/home/products/ortho-vision-max-analyzer
- 25. Dolci A, Giavarina D, Pasqualetti S, Szoke D, Panteghini M. Total laboratory automation: Do stat tests still matter? Clin Biochem. 2017;50: 605–611. pmid:28390779
- 26. Chung HJ, Song YK, Hwang SH, Lee DH, Sugiura T. Experimental fusion of different versions of the total laboratory automation system and improvement of laboratory turnaround time. J Clin Lab Anal. 2018;32: e22400. pmid:29479855
- 27. Lippi G, Da Rin G. Advantages and limitations of total laboratory automation: a personal overview. Clin Chem Lab Med. 2019;57: 802–811. pmid:30710480
- 28. South SF, Casina TS, Li L. Exponential error reduction in pretransfusion testing with automation. Transfusion. 2012;52: 81S–87S. pmid:22882101
- 29. Bhagwat SN, Sharma JH, Jose J, Modi CJ. Comparison between conventional and automated techniques for blood grouping and crossmatching: Experience from a tertiary care centre. J Lab Physicians. 2015;7: 96–102. pmid:26417159
- 30. Koh YE, Yoon J, Kwon SH, Kim YH, Choi JY, Kim JY, et al. Evaluation of the automated blood bank instrument QWALYS-3 for cross-matching tests. Korean J Blood Transfus. 2014;25: 218–225.
- 31.
Korea enforcement decree of the bioethics and safety act. Article 36 (2). 2012. [cited 2019 May 20]. Database: National Law Information Center. Available from: http://www.law.go.kr.
- 32. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig L, et al. STARD 2015: An updated list of essential items for reporting diagnostic accuracy studies. Clin Chem. 2015;61: 1446–1452. pmid:26510957
- 33.
Machin D, Campbell MJ, Tan SB, Tan SH. Sample sizes for clinical, laboratory and epidemiology studies. 4th ed. Hoboken, Wiley-Blackwell; 2018.
- 34.
Indrayan A, Malhotra RK. Medical Biostatistics. 4th ed. Florida, Taylor & Frnacis; 2018.
- 35.
International Electrotechnical Commission. International Standard 60812 2018, Failure modes and effects analysis (FMEA and FMECA), 3rd ed. Geneva: International Electrotechnical Commission; 2018.
- 36. Shin KH, Kim HH, Chang CL, Lee EY. Economic and workflow analysis of a blood bank automated system. Ann Lab Med. 2013;33: 268–273. pmid:23826563
- 37. Sim J, Wright CC. The kappa statistic in reliability studies: Use, interpretation, and sample size requirements. Phys Ther. 2005;85: 257–268. pmid:15733050
- 38.
Balenton N. Nonparametric statistics. In: Khakshooy AM, Chiappelli F, editors. Practical biostatistics in translational healthcare. Germany: Springer; 2018. pp. 123–126.
- 39. Burckhardt I, Last K, Zimmermann S. Shorter incubation times for detecting multi-drug resistant bacteria in patient samples: Defining early imaging time points using growth kinetics and total laboratory automation. Ann Lab Med. 2019;39: 43–49. pmid:30215229
- 40. Choi Q, Kim HJ, Kim JW, Kwon GC, Koo SH. Manual versus automated streaking system in clinical microbiology laboratory: Performance evaluation of Previ Isola for blood culture and body fluid samples. J Clin Lab Anal. 2018;32: e22373. pmid:29314254
- 41. Li Z, Han R, Yan Z, Li L, Feng Z. Antinuclear antibodies detection: A comparative study between automated recognition and conventional visual interpretation. J Clin Lab Anal. 2019;33: e22619. pmid:30030865
- 42. Chaudhary R, Agarwal N. Safety of type and screen method compared to conventional antiglobulin crossmatch procedures for compatibility testing in Indian setting. Asian J Transfus Sci. 2011;5: 157–159. pmid:21897596
- 43. Alavi-Moghaddam M, Bardeh M, Alimohammadi H, Emami H, Hosseini-Zijoud SM. Blood transfusion practice before and after implementation of type and screen protocol in emergency department of a university affiliated hospital in Iran. Emerg Med Int. 2014;2014: 316463. pmid:25254117
- 44. Kumari S. Blood transfusion practices in a tertiary care center in Northern India. J Lab Physicians. 2017;9: 71–75. pmid:28367018